Cable television (CATV), originally introduced in the late 1940's as a way to transmit television signals by coaxial cables to houses in areas of poor reception, has over the years been modified and extended to enable the cable medium to transport a growing number of different types of digital data, including both digital television and broadband Internet data.
One of the most significant improvements occurred in the 1990's, when a number of major electronics and cable operator companies, working through CableLabs, a non-profit R&D consortium, introduced the Data Over Cable Service Interface Specification (DOCSIS). First introduced in the late 1990's as DOCSIS version 1.0, and upgraded many times since (currently at DOCSIS version 3.0), the DOCSIS standard defines the Physical Layers (PHY) and Media Access Control (MAC) layers needed to send relatively large amounts of digital data through coaxial cables that were originally designed to handle analog standard definition television channels.
Originally, analog television (in the US) transmitted television channels as a series of roughly 6 MHz bandwidth radiofrequency waveforms at frequencies ranging from about 54 MHz (originally used for VHF Channel 2) up to about 885 MHz for now no-longer used UHF channel 83. This television signal was transmitted as a combination amplitude modulated signal (for the black and white portion), quadrature-amplitude modulated signal (for the color portion), and frequency modulated signal (for the audio portion), and this combined signal will be designated as a Frequency Division Multiplexed (FDM) signal.
With the advent of digital television and high definition television standardization in the late 1980's and early 1990's, the basic 6 MHz bandwidth spectrum of analog television was retained, but the modulation scheme was changed to a more sophisticated and higher data rate Quadrature Amplitude Modulation (QAM) scheme, which can encode digital information onto a very complex QAM analog signal (waveform).
The DOCSIS standard built upon this analog and digital TV foundation, and specified additional standards to provide broadband Internet services (Internet protocols, or IP), voice over IP, custom video on demand, and other modern services based upon the QAM data transmission waveforms (generally also 6 MHz wide) previously established for digital and high definition television.
As a result, by a series of steps, simple coaxial cables, originally run at great expense to millions of households starting from the 1950's and 1960's, has been gradually upgraded to accommodate ever increasing demands for digital data. At each house (or apartment, office, store, restaurant or other location), the household connects to the CATV cable by a cable modem, uses the cable modem to extract downstream DOCSIS digital data (frequently used for high-speed Internet), and inject upstream DOCSIS digital data (again frequently used for high-speed Internet applications).
Unfortunately, even in a coax cable, there is a finite amount of bandwidth available to transmit data. Coax cables and their associated radiofrequency interface equipment have typically only used the frequency range under about 1000 MHz, and so there are limits to how much data the 1950's era coaxial cable can ultimately transmit.
By contrast, optical fiber (fiber optics, fiber) technology, which uses much higher optical frequencies (with wavelengths typically in the 800-2000 nanometer range), can transmit a much higher amount of data. Optical fiber data rates typically are in the tens or even hundreds of gigabits per second. Indeed, the entire RF CATV cable spectrum from 0 to 1000 MHz can be converted to optical wavelengths (such as 1310 nm or 1550 nm), be carried over an optical fiber, and then be converted back to the full RF CATV cable spectrum at the other end of the fiber, without coming close to exhausting the ability of the optical fiber to carry additional data.
This conversion process can be achieved by relatively simple optical to digital or digital to optical converters, in which the CATV RF waveforms are simply converted back and forth to a light signal by simple (“dumb”) E/O or 0/E converters, located in nodes that connect optical fibers to CATV cable (fiber nodes).
The higher data carrying capacity of optical fibers allows additional data to be carried as well, and in some schemes, the essentially analog (digital encoded in analog) spectrum of CATV waveforms is carried at one optical wavelength (such as 1310 nm), and digital data encoded by entirely different protocols may be carried at an alternate optical wavelength (such as 1550 nm). This dual scheme is often referred to as wavelength-division multiplexing.
Optical fiber technology has been widely used for high capacity computer networks, and these networks often do not use the DOCSIS protocols or QAM protocols to transmit data. Rather, these high capacity computer networks often use entirely different types of data transmission protocols, such as the Ethernet protocols IEEE 802.3ah, 1000BASE-LX10, 1000Base-BX10, and others. These networks and protocols are often referred to as GigE networks, which is an abbreviation of the Gigabyte speeds and Ethernet protocols used for fiber based computer network.
Thus if a user desires to transfer computer data from RF QAM waveforms transported over a CATV cable to a high speed GigE fiber network, the data must be transformed back and forth between the DOCSIS cable QAM waveforms and the alternate protocols (often Ethernet protocols) used in fiber GigE networks.
Although ideally, the best way to satisfy the ever increasing household demand for digital data (e.g. video—on demand, high speed Internet, voice over IP, etc.) would be by extending optical fiber to each household, this would be an incredibly expensive solution. By contrast, cable based CATV solutions have already been implemented for tens of millions of households, and this expense has already been borne and amortized over decades of use, starting from the 1950s. As a result, it is far more economically attractive to find schemes enable the existing, if bandwidth limited, CATV cable system, to be further extended to meet the ever growing demands for additional data.
Cable System Components:
At the plant or “head” end of a typical CATV cable network (cable), the challenging task of combining the many different types of data (analog television channels, digital television channels, on-demand channels, voice over IP, DOCSIS channels, etc.) and sending this data to users (households) scattered through many different neighborhoods in various regions of towns, cities, counties and even states is handled, in part, by Cable Modem Termination Systems (CMTS) devices. These CMTS devices connect to the various data sources (television stations, video servers, the Internet, etc.) at one end, and to many different CATV cables at the other end.
Typically the CMTS device will have a connection to the various data sources and appropriate data switches (such as a Level 2/3 switch) at one end, and often a plurality of different line cards (often physically packaged to look like blade servers, and put into a main CTMS box that holds multiple line cards) at the other end. Each line card will typically be connected to either cables or optical fibers that travel away from the cable head towards various groups of multiple neighborhoods, where typically each group of multiple neighborhoods will be in a roughly contiguous geographic region. The line card cables or optical fibers are then typically subdivided further by various splitters and nodes, and eventually the signals flow to the individual neighborhoods, each served by its own CATV cable.
At the neighborhood level, an individual CATV cable will serve between about 25 and a few hundred households (houses, apartments). These connect to the individual cable by cable modems. Here each cable modem will be considered to be a household or “house”, regardless of if the cable modem serves a house, apartment, office, workplace, or other application.
The CMTS line cards will typically contain at least the MAC and PHY devices needed to transmit and receive the appropriate CATV signals. Typically the line card PHY devices will contain a plurality of QAM modulators that can modulate the digital signals that a Level 2/3 switch has sent to that particular line card, and send the signals out over cable or fiber as a plurality of QAM channels. The line cards will also typically contain MAC and PHY devices to receive upstream data sent back to the cable head from the various cables and cable modems in the field.
It is impractical to directly connect each individual neighborhood CATV cable directly to the cable plant. Rather cable networks are arranged in more complex schemes, where the signals to and from many different individual neighborhoods are combined by the network prior to reaching the cable plant or cable head. Thus each CMTS line card will typically send and receive signals to and from multiple neighborhoods.
Instead of sending and receiving data by cable, the various CMTS line cards can instead communicate to their various groups of neighborhoods by optical fiber. However it is also impractical to run individual fibers directly from individual neighborhoods to the cable plant or cable head as well. Thus fiber networks are also usually arranged in more complex schemes, where the signals to and from different individual neighborhoods are also combined by the optical fiber network before the signals reach the cable plant or cable head.
At a minimum, the optical fiber network will at least typically split (or combine) the fiber signals, often by “dumb” optical fiber splitters/combiners (here called splitters) that do not alter the fiber signal, and the split signal then will be sent by sub-fibers to the various neighborhoods. There, the optical fiber signal can be converted to and from a RF signal (suitable for the individual cable) by a “dumb” fiber node that itself simply converts the optical to RF and RF to optical signals without otherwise altering their content. These hybrid optical fiber to cable networks are called Hybrid Fiber Cable (HFC) networks.
Prior art work with various types of CMTS systems and fiber nodes includes Liva et. al., U.S. Pat. No. 7,149,223; Sucharczuk et. al. US patent application 2007/0189770; and Amit, U.S. Pat. No. 7,197,045.
Typically, nearly all CATV users want immediate access to at least a standard set of cable television channels, and thus to satisfy this basic expectation, usually all CATV cables will receive a basic set of television channels that correspond to this “basic” or “standard” package (which may include various commonly used premium channels as well). Additionally, most users will wish access to a wide range of individualized data, and here the limited bandwidth of the CATV cable starts to become more of a nuisance.
As a first step towards more efficient cable utilization, analog television is being phased out, freeing much FDM bandwidth (analog standard definition TV channels) that can be replaced by more efficient QAM channels carrying both digital TV and DOCSIS data. However phasing out old-fashioned FDM TV signals, although freeing up additional cable bandwidth, will at most satisfy the ever increasing household demand for digital TV and DOCSIS services (data) for only a few years. Thus additional methods to supply a greater amount of data, in particular on-demand video data, voice over IP data, broadband Internet (IP) data, and other data, are desirable.
DOCSIS Standards:
Unless otherwise specified references herein to “DOCSIS” will refer to the Cablelabs DOCSIS® 3.0 specifications. These are more specifically defined in the following publications: Data-Over-Cable Service Interface Specifications DOCSIS 3.0 Security Specification CM-SP-SECv3.0-113-100611; Cable Modem to Customer Premise Equipment Interface Specification CM-SP-CMCIv3.0-101-080320; Physical Layer Specification CM-SP-PHYv3.0-110-111117; MAC and Upper Layer Protocols Interface Specification CM-SP-MULPIv3.0418-120329; Operations Support System Interface Specification CM-SP-OSSIv3.0418-120329. Additional documentation can be found in the DOCSIS 3.0 Technical Reports CM-TR-MGMTv3.0-DIFF-V01-071228; and CM-TR-OSSIv3.0-CM-V01-080926. For purposes of this specification, features that implement an otherwise compatible subset of the DOCSIS 3.0 specification are termed a DOCSIS subset, and features that implement either additional functions not specified in DOCSIS 3.0, or incompatible with DOCSIS 3.0, are termed “non-DOCSIS functionality”.
Remotely situated QAM modulators:
Liva et. al., in U.S. Pat. No. 6,933,016 taught a method of transmitting an information channel by a unique method of processing the information channel, transmitting the information channel to a destination by packet techniques, and then reconstructing the channel. Additionally Sawyer, in US Publication 2003/0066087, taught a hybrid distributed cable modem termination system having mini fiber nodes containing CMTS modulators remotely located from the head end.
Field-Programmable Gate Array (FPGA) Technology:
Field-programmable gate arrays (FPGA), a type of programmable logic device (PLD), are integrated circuit devices and “chips” designed to allow the configuration of the chip's various internal electrical circuits to be reconfigured after the chip has been manufactured. FPGAs contain programmable logic blocks with reconfigurable connections that allow the wiring between the various logic gates in the blocks to be rewired, even after the chip has been incorporated into other devices. In addition to digital functions, FPGA can handle analog functions. Various mixed signal FPGA, with integrated analog to digital converters (ADC) and digital to analog converters (DAC) are also available. Examples of FPGA include the popular Artix, Kintex, Virtex, and Spartan series of chips produced by Xilinx Inc., San Jose, Calif., the popular Cyclone, Arria, Stratix series of chips produced by Altera Corporation, San Jose Calif., and others.
Digital Signal Processor (DSP) Technology:
Digital signal processor (DSP) devices and “chips” are microprocessors with an architecture that is specialized for high speed digital signal processing. Although standard processors can perform complex signal processing as well, due to the nature of the standard instruction set hardware, complex signal processing often requires a large (hundreds, thousands, or more) number of instructions to perform complex functions. By contrast, DSP chips often contain at least some specialized hardware for digital signal processing, including circular buffers, separate program and data memories (e.g. Harvard architecture), very long instruction words (VLIW), various types of single instruction multiple data (SIMD) instructions, fast multiply-accumulate (MAC) hardware, bit reversed addressing, special loop controls, and the like. This specialized hardware allow complex signal processing to be done in a relatively few number of operations, thus often speeding up complex computations by many orders of magnitude in time. To further reduce processing time DSP are often constructed without memory management units, thus avoiding time delays due to memory management unit induced context switching.
Examples of DSP include the popular C6000 series of DSP produced by Texas Instruments, Inc, Dallas Tex., the StarCore DSP produced by Freescale Semiconductor Holdings, Ltd., Austin Tex., and others.
Examples of the use of FPGA and DSP to produce dynamically reconfigurable communications devices include Dick, U.S. Pat. No. 7,583,725, and Raha et. al., U.S. Pat. No. 7,724,815, both assigned to Xilinx, the contents of both of which are incorporated herein by reference.